Genome wide DNA sequencing is now being adopted in clinical practice and an increasing number of variants are identified in epilepsy-associated genes, yet the clinical interpretation of the new variants is challenging. Some of the variants are known to be either pathological or benign, yet a majority of the gene variations remain unknown for their functional consequence. A large number of Variant of Uncertain Significance (VUS) are becoming commonplace in genes for human diseases, providing a significant barrier in making diagnoses and implementing therapies. Bioinformatic approaches can provide some insight into pathogenic probability of VUS alleles, but functional studies in animal model systems are often needed to make definitive of pathogenicity assignments. The expense and long timelines of mouse model production make the use of alternative small animal models attractive. In this proposal, the ?C. elegans?nematode is used as an alternative model capable of fast high-throughput production and screening. Human genes are installed as gene-swap replacements of the native disease-gene homologs. In preliminary work, gene-swap humanization of STXBP1 in the ?unc-18?locus rescued severe locomotion and behavior defects present in the gene knock-out animals. Pathogenic variants into the STXBP1??gene-swap loci leads to significant disruption of activity. In this proposal, significant and novel improvements are made to our existing pipeline for the functional analysis of variants in vivo. In Aim 1, the relevance and extensibility of the C. elegans model system for studying human disease is improved through simultaneous humanization of multiple related loci. In Aim 2, new methods are developed for molecular phenotyping, improving the resolution of inputs pathogenicity determination algorithms, and yielding mechanism-of-action level readouts to variant manipulations. In Aim 3, the improvements to the pipeline are tested to quantify gains in pathogenicity determination on a test set of variants.